We’re now far along into the narrative on how data will “change the way we live and work.” Further, that narrative is beginning to develop into more than an abstract promise, as we see examples and case studies chronicled regularly in such sources as The Harvard Business Review and the Stanford Social Innovation Review. But, we have a long way to go – and one particularly nagging structural problem to address in the way we ensure the flow of high quality data as we aim to do good.
We are able to optimize, iterate, and experiment in ways that we could not imagine even ten years ago. The first to see these effects were internet businesses where the ability to capture, store, and organize data is now cheap and easy. And these data have slowly made their way offline and into every aspect of our lives. We have to put ourselves in position to take manage the trend on a local level as well as on a sector level.
While I was Director of Research & Analytics at a large youth services organization in Chicago I helped found a collaborative with four similar organizations. We shared our enrollment information and then merged that with the academic records of the students we were serving. We were then able to see how each program and organization was affecting outcomes like school day attendance, grades, and disciplinary records. These learnings were shared across the group and we were all able to improve our programs based upon what we were learning from each other.
On the whole, however, the social sector, hasn’t yet seen the kind of transformation that is taking place in other sectors, the kind that is possible when we improve our data sharing. One big reason is a lingering silo mentality. The silos themselves (the operating areas) may not be the problem, but freezing the data within them is definitely a problem. We lack, as a discipline, the structured and regular sharing of data across and within organizations whether for real-time dashboards, analysis, business modeling, or simply to ask new questions. In your organization, does your fundraising data talk to your marketing data? Does your marketing data to your program data?
This is a problem because the people we work with are often in multiple systems. The same individual gives you money, receives your emails, and sometimes even participates in your programs. Yet you cannot see this entire person; you can only see slices of them. This hinders your ability to optimize, iterate, and experiment.
Think of it this way. Would you rather have to create the best 5 card poker hand with the first 5 cards you were dealt or all 52? With the first 5 cards you might get a high pair but with all 52 at your disposal you are guaranteed a royal flush.
By leaving our data in silos we are playing with the first 5 cards we are dealt. By bringing these systems together we are expanding the number of cards at our disposal and our insights get more and more valuable.
As I mentioned, silos don’t just operate within our organizations, they operate between them as well. Because nonprofits don’t often achieve the scale of for-profits, no single nonprofit would ever achieve the quantity of data that a Google has. Imagine if there was a Google in every city, and the data collected in one was never shared with another. Google wouldn’t be nearly as successful and it certainly wouldn’t have discovered some of the insights it has.
This is what exists in the nonprofit sector on an even larger scale. What needs to happen is for nonprofits to join forces and appropriately share their data with one another for the benefit of all.
The current data movement has the potential to revolutionize the nonprofit sector in a way it didn’t even do for businesses. As we are able to more easily track and measure the outcomes that nonprofits produce we will, for the first time, find ourselves accountable not just fiscally but for the quality of the outcomes we are delivering. The organizations that will thrive in this future are those that have a head start and have overcome the myriad of silos that exist in the sector.
…
Your turn.
[poll id=’24’]